Popular repositories Loading

1. flashdec flashdec Public

   Single-GPU LLM decode research prototype: paged KV cache, Triton attention, CUDA append, scheduling, shared prefixes, and multi-layer transactions.

   Python 1

2. xiaosong_home xiaosong_home Public

3. gpgpu_test gpgpu_test Public

   Python

4. QwenServe-12G QwenServe-12G Public

   Reproducible vLLM inference optimization and QLoRA experiments on an RTX 5070 12GB GPU

   Python